A system and method are disclosed for monitoring usage of resources (e.g., hosts, instances, applications, etc.) in a datacenter. Customers, developers and system administrators can collect and track metrics, gain insight, and react to keep applications and businesses running smoothly by providing system-wide visibility into resource utilization, application performance, and operational health. Users can programmatically retrieve monitoring data and view heat maps to assist in troubleshooting, spotting trends, and taking automated action based on the state of a cloud environment. Users can further monitor resources in real-time, so that metrics such as CPU utilization, latency, memory usage, transaction volumes, error rates, etc. can be visualized.